Cocaine dependence (CocDep) is a significant public health problem with rising rates in young adults. Evidence suggests that chronic cocaine use is associated with changes in brain structure and cognitive impairments that may complicate treatment and recovery. Research using diffusion tensor imaging (DTI) suggests that white matter (WM) undergoes substantial changes in CocDep as a result of chronic use and soon after drug use ceases. Studies in active cocaine users have reported that fractional anisotropy (FA), a measure of WM integrity, is lower in the inferior frontal WM and corpus callosum. However, when we studied former cocaine-dependent patients who remained abstinent, we found that FA in inferior frontal WM was elevated (compared with nonusers) in the early abstinence period, perhaps due to rapidly-evolving neuroadaptations. In our pilot study, elevations in frontal WM FA in CocDep were associated with cognitive impairment, correlated with longer durations of prior drug use and were attenuated by longer periods of abstinence. Frontal FA was highest in patients scanned in early abstinence and showed significant and widespread recovery over the first three to six months of abstinence in subjects scanned longitudinally. Subjects that did not remain in treatment for six months had higher FA in the inferior frontal WM and smaller caudate nuclei than did subjects that remained in treatment. We hypothesize that a biphasic response occurs in FA in early abstinence (an initial, rapid increase, followed by a slower decline) that is related to structural WM instability. Since treatment retention in early abstinence is critical to long-term outcomes, understanding the continuum of WM changes during this time has important implications for optimizing treatment. We hypothesize that disturbances in the integrity of mesolimbic, mesocortical and cortico-striatal WM pathways represent a crucial defect in CocDep that are related to clinical status and treatment retention. This project will address both of these primary hypotheses by longitudinally testing cocaine-dependent patients in long term rehabilitation treatment using MRI and cognitive testing and comparing them with longitudinally-tested non- users. 160 patients will be initially tested within the first 10 days following drug cessation and approximately 100 patients will be tested again at 5 weeks abstinence to determine whether WM FA increases during the early abstinence period. 6- and 12-month follow-up sessions will permit characterization of the longitudinal changes in WM measures in patients that remain in treatment. This design also will allow us to better understand the neural factors that contribute to treatment retention and dropout. Patients also will be followed in the community to provide information about neural factors contributing to relapse and ongoing abstinence. It is hoped that such information will inform improvements in the treatment of CocDep.Project Narrative Cocaine dependence is a significant public health problem that remains difficult to treat. This research will increase our understanding of the brain changes associated with abstinence from cocaine and treatment- related measures. Thus, it will provide new information about cocaine addiction and recovery that may improve the medical management of cocaine dependence, if not addictions in general.